PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans

Abstract Delays or misdiagnoses in detecting pneumoperitoneum can significantly increase mortality and morbidity. We developed and validated a deep learning model designed to identify pneumoperitoneum in computed tomography images. The model is trained on abdominal scans from Far Eastern Memorial Ho...

Full description

Saved in:
Bibliographic Details
Main Authors: I-Min Chiu, Teng-Yi Huang, David Ouyang, Wei-Che Lin, Yi-Ju Pan, Chia-Yin Lu, Kuei-Hong Kuo
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-024-54043-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850196546018082816
author I-Min Chiu
Teng-Yi Huang
David Ouyang
Wei-Che Lin
Yi-Ju Pan
Chia-Yin Lu
Kuei-Hong Kuo
author_facet I-Min Chiu
Teng-Yi Huang
David Ouyang
Wei-Che Lin
Yi-Ju Pan
Chia-Yin Lu
Kuei-Hong Kuo
author_sort I-Min Chiu
collection DOAJ
description Abstract Delays or misdiagnoses in detecting pneumoperitoneum can significantly increase mortality and morbidity. We developed and validated a deep learning model designed to identify pneumoperitoneum in computed tomography images. The model is trained on abdominal scans from Far Eastern Memorial Hospital (January 2012–December 2021) and evaluated using a simulated test set (14,039 scans) and a prospective test set (6351 scans) collected from the same center between December 2022 and May 2023. External validation included 480 scans from Cedars-Sinai Medical Center. Overall, the model achieves a sensitivity of 0.81–0.83 and a specificity of 0.97–0.99 across retrospective, prospective, and external validation; sensitivity improves to 0.92–0.98 when cases with a small amount of free air (total volume <10 ml) are excluded. These findings suggest that the model can deliver accurate and consistent predictions for pneumoperitoneum in computed tomography scans with segmented masks, potentially accelerating diagnostic and treatment workflows in emergency care.
format Article
id doaj-art-af2ca42896254049aeeb4cd1e9902be3
institution OA Journals
issn 2041-1723
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-af2ca42896254049aeeb4cd1e9902be32025-08-20T02:13:26ZengNature PortfolioNature Communications2041-17232024-11-011511710.1038/s41467-024-54043-1PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scansI-Min Chiu0Teng-Yi Huang1David Ouyang2Wei-Che Lin3Yi-Ju Pan4Chia-Yin Lu5Kuei-Hong Kuo6Department of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical CenterDepartment of Electrical Engineering, National Taiwan University of Science and TechnologyDepartment of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical CenterDepartment of Diagnostic Radiology, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of MedicineDepartment of Psychiatry, Far Eastern Memorial HospitalDepartment of Diagnostic Radiology, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of MedicineDivision of Medical Image, Far Eastern Memorial HospitalAbstract Delays or misdiagnoses in detecting pneumoperitoneum can significantly increase mortality and morbidity. We developed and validated a deep learning model designed to identify pneumoperitoneum in computed tomography images. The model is trained on abdominal scans from Far Eastern Memorial Hospital (January 2012–December 2021) and evaluated using a simulated test set (14,039 scans) and a prospective test set (6351 scans) collected from the same center between December 2022 and May 2023. External validation included 480 scans from Cedars-Sinai Medical Center. Overall, the model achieves a sensitivity of 0.81–0.83 and a specificity of 0.97–0.99 across retrospective, prospective, and external validation; sensitivity improves to 0.92–0.98 when cases with a small amount of free air (total volume <10 ml) are excluded. These findings suggest that the model can deliver accurate and consistent predictions for pneumoperitoneum in computed tomography scans with segmented masks, potentially accelerating diagnostic and treatment workflows in emergency care.https://doi.org/10.1038/s41467-024-54043-1
spellingShingle I-Min Chiu
Teng-Yi Huang
David Ouyang
Wei-Che Lin
Yi-Ju Pan
Chia-Yin Lu
Kuei-Hong Kuo
PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
Nature Communications
title PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
title_full PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
title_fullStr PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
title_full_unstemmed PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
title_short PACT-3D, a deep learning algorithm for pneumoperitoneum detection in abdominal CT scans
title_sort pact 3d a deep learning algorithm for pneumoperitoneum detection in abdominal ct scans
url https://doi.org/10.1038/s41467-024-54043-1
work_keys_str_mv AT iminchiu pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT tengyihuang pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT davidouyang pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT weichelin pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT yijupan pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT chiayinlu pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans
AT kueihongkuo pact3dadeeplearningalgorithmforpneumoperitoneumdetectioninabdominalctscans